THE RECENT REMARKABLE SUNSETS. 
179 
the reflecting surface at that elevation, and this gives 4042 
miles as the distance from the centre of the earth, or 42 
miles higher than the surface of the earth, or say 40 miles as 
the height of the reflecting surface in the atmosphere after 
allowing for the disturbing effect of refraction in reducing the 
apparent depression of the sun below the horizon. This 40 
miles was named by Helmholtz, junr., as the approximate 
height, calculated from observations at Berlin of the special 
sunset phenomena. 
The theory of an unusual amount of aqueous vapour in 
the atmosphere as a cause for the phenomenon is seen to be 
untenable when this great height of 40 miles is recognised, 
because the limit of vapour in the atmosphere capable of 
forming clouds is much below that height, as the rarefaction 
of the atmosphere at that height may be considered much too 
great to allow of aqueous vapour being held there in the 
atmosphere of sufficient density to allow visible cloud to be 
formed or strata capable of giving a sunset reflection. 
In the case of the other theory of diffused meteoric 
matter in the atmosphere as the cause of the phenomenon, it 
has to be noted that this matter would necessarily be diffused 
with an approximate uniformity over the whole atmosphere 
surrounding the earth, in the event of the earth meeting 
with and plunging into any collection of meteoric matter in 
the course of its orbit; this collection of matter being in 
front of the earth in its course, and the earth turning round 
completely in twenty-four hours, would result in a general 
distribution of the meteoric matter over its whole surface. 
But the consequence of this would be a simultaneous appear¬ 
ance of the special phenomenon in different parts of the 
earth, and this is directly opposed to the observed facts that 
the phenomenon was distinctly local, and travelled over the 
surface of the earth with great uniformity of motion. There 
were two different directions and rates of motion: first a 
direct westerly translation near the line of the equator that 
was traced three-quarters of the distance round the earth at 
a very uniform speed of about 70 miles an hour; and secondly 
a much slower diffusion northwards and southwards, reaching 
30° from the equator in about a month, 50° in three months, 
and G0° in four months. In the chart given in the former 
paper upon this subject (“Midland Naturalist,” Plate I., 
January, 1884), the westerly travel of the phenomenon is 
shown at 12,000 miles distance, or halfway round the earth 
in a week; and since that time information has been received 
of its reaching the Sandwich Islands at 18,000 miles distance, 
or three-quarters round the earth, in ten days, both observa- 
